INDEX

Explanations

references to locks and locking mechanisms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

security locks and latches

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 phenomenon

-0.35

misu

-0.33

TypedDataSet

-0.33

republik

-0.33

ália

-0.33

 effusion

-0.33

 phenomena

-0.32

 appreciation

-0.32

 appropriation

-0.32

 cadre

-0.32

POSITIVE LOGITS

 security

0.59

 locksmith

0.58

 SECURITY

0.56

security

0.54

 замок

0.53

 lock

0.52

Security

0.52

 Security

0.52

 padlock

0.51

 locks

0.51

Activations Density 0.009%

references to locks and locking mechanisms

security locks and latches

No Comments

No Known Activations